Can the migration script run without a checkout of the repository? #537
Labels
enhancement
Improvements to an existing feature
Migration
Relating to the RACK version migration script
When we distribute the migration tool, we do so in a copy of the directory that is not a check out of the git repository.
Yet, to run, the migration script currently depends on being in a git repository, as it tries to run git commands to, for instance, resolve tags to commit IDs.
While the
rack_crawl
tool definitely needs git to work, it seems like we could pre-compute all relevant data so thatrack_migrate
no longer needs to be in a git repo. After all, we already have most of the commits data.I think it is possible to cut this dependency, but I may be forgetting some reason why git access is necessary. I will investigate this possibility soon.
(This was prompted by a request from a user at Boeing.)
The text was updated successfully, but these errors were encountered: